Treating Depression in Disadvantaged Women: What is the evidence? Megan Dwight Johnson, MD MPH Associate Professor Medical Director, UWMC Inpatient Psychiatry Department of Psychiatry and Behavioral Sciences University of Washington
Overview Can antidepressants effectively treat depression in disadvantaged populations? Can psychotherapy effectively treat depression in disadvantaged populations? How can practices and providers improve rates and quality of care?
What is Major Depression? NOT having a bad day, a bad attitude, or normal sadness A pervasive depressed mood or lack of interest/pleasure for more than 2 weeks, plus Impaired sleep Changes in appetite Decreased energy Poor concentration Guilty ruminations Psychomotor changes Thoughts of death/suicide
What is Dysthymia? Depressed mood most of the day, more days than not for > 2 years >2 of the following: Change in appetite Change in sleep Low energy Low self-esteem Poor concentration Feelings of hopelessness
Epidemiology of Depression Lifetime prevalence Major Depression: 10-18% Dysthymia: 2-5% National Comorbidity Study-Replication Although there is variation between studies, rates appear similar across race and ethnic groups controlling for socioeconomic status Low income persons are at increased risk for depression (including post-partum depression) (Kessler et al 1994) Among Latinos, recent immigrants may have lower rates of depression compared to US born
Low rates of depression care in lowincome and ethnic minority groups Of African Americans with a diagnosable mental health condition: Only 33% of African Americans received any treatment Only 16% of African Americans received specialty mental health care (Kessler et al 1994) Among Latinos with diagnosable mental health condition: Fewer than 1 in 5 contact a general health provider (<1 in 10 among recent immigrants) Fewer than 1 in 11 contact a mental health specialist (<1 in 20 among recent immigrants)
Less is known about rates of care for Asian Americans and Native Americans Asian Americans Diverse group not well represented in epidemiologic studies Lower rates of insurance than whites Lower use of mental health services than whites Severity of illness greater when services are sought Native Americans Data extremely sparse Lower rates of insurance than whites
Quality of Depression Care Household survey of 1636 adults Rates of appropriate care for depressive and anxiety disorders: 34% for whites 24% for Latinos 17% for African Americans» Young et al 2001
Poorer quality of depression care for persons of color Even when primary care providers diagnose depression and recommend treatment: African Americans (OR=0.30) and Latinos (OR=0.42) are less likely than whites to report taking an antidepressant Latinos are less likely than whites to obtain specialty mental health services (OR=0.50) (Miranda & Cooper 2004) In one primary care study,12.8% of Latinos received appropriate depression care compared to 35.3% of whites (Miranda et al 2003)
Evidence-Based Treatment for Depressive Disorders Antidepressant Medication Psychotherapy/Behavioral interventions
Antidepressant Medications Selective Serotonin Re-uptake Inhibitors (SSRIs) Fluoxetine, sertraline, paroxetine, citalopram, escitalopram, fluvoxamine Other antidepressants Venlafaxine, duloxetine, bupropion Tricyclic antidepressants Imipramine, amitriptyline, desipramine, nortriptyline All shown effective in randomized trials
Low-income and minority patients are underrepresented in treatment trials Only 7% of subjects participating in depression trials are identified as racial/ethnic minorities (Surgeon General s Report) Recent initiatives from the National Institute of Mental Health aimed at increasing participation of minorities in clinical trials
Antidepressant treatment Drug metabolism is in part genetically determined Distribution of genetic polymorphisms varies across ethnic groups Genetic differences may interact with differences in diet and use of alternative medicine Thus, efficacy and sensitivity may vary between ethnic groups
Cytochrome P-450 Affect metabolism of most antidepressants Alleles for inactive cytochrome p-450 Heterozygous: intermediate metabolizers Homozygous: poor metabolizers Rates of poor metabolizers highest in Asians compared to other ethnic groups May lead to increased side effects at usual doses Rates of ultra-high metabolizers highest in Arab and Ethiopian populations May lead to low blood levels and decreased efficacy at usual doses
Paroxetine in Minority Patients Pooled analysis of the large paroxetine clinical trials database No consistent differences in response and remission rates for both depression and various anxiety disorders similar in Latino, African American, Asian, and Caucasians Results likely similar for all the SSRI s Roy-Byrne et al 2005
Antidepressants in Ethnic Minorities Despite under representation of ethnic minorities in research studies Despite possible genetic variation in medication effects Existing evidence suggests that standard antidepressants are effective in ethnic minority patients
Evidence-Based Psychotherapies Cognitive Behavioral Therapy Behavioral Activation Interpersonal Therapy Problem Solving Therapy Structured May be manualized
Cognitive Behavioral Therapy Depression results from a cycle of: Events Negative Cognitions Mood CBT helps clients to: Identify and correct negative cognitions Increasing positive interpersonal interactions and pleasurable activities Individual or group 6 randomized trials showing equal efficacy with medication for depression
Behavioral Activation As effective as CBT Shown effective as medication in severe depression May require less training to deliver than CBT
Interpersonal Therapy A structured, supportive therapy linking recent interpersonal events to mood or other problems, paying systematic attention to current personal relationships, life transitions, role conflicts and losses
Problem Solving Therapy Involves training individuals: problem orientation problem definition and formulation generation of alternatives decision making solution implementation and verification Shown effective in primary care settings
Adapting psychotherapy for disadvantaged patients Language (translation, reading level) Address cultural similarities/differences between client and provider Goals should be culturally compatible The conceptualization of the problem and treatment should be culturally congruent Manual content (i.e. vignettes) should be culturally appropriate Activities should be economically and culturally appropriate Take into consideration the sociocultural context Extensive outreach Flexible schedule
WE-CARE: Miranda et al 2003 267 Women with Major Depression 44% African American 50% Latino 60% below federal poverty level Recruited from WIC and Family Planning offices Randomized to either: Cognitive Behavioral Therapy (8 weeks) Antidepressants (Paroxetine or Bupropion) Community Referral
WE-CARE Outcomes Medication (p<0.001) and CBT (p=.006) more likely to improve depression symptoms (HDRS) than community referral Rates of remission (HDRS<7): 44.4% with medication 32.2% with CBT 28.1% with community referral
WE-CARE 1-Year Outcomes 16 14 Hamilton Depression Rating Scale Score 12 10 8 6 4 Community Referral Cognitive Behavioral Therapy Medication 2 0 30 60 90 120 150 180 210 240 270 300 330 360 Days Since Baseline Miranda et al, J Con Clin Psychol, 2006
WE-CARE Process of Care Of those assigned to medication: 75% received 9 weeks of guideline level antidepressants 45% received 24 weeks of guideline level antidepressants Average of 8 contacts prior to first medication visit Of those assigned to CBT: 53% received 4 or more sessions 36% received 6 or more sessions Average of 10 contacts prior to first session
WE-CARE Patients with Depression and PTSD At baseline, 33% of patients had comorbid PTSD Compared to women with depression alone, those with PTSD and depression: Had higher exposure to assualtive violence Had higher levels of depression/anxiety Had more functional impairment At 1 year follow-up: Women with and without PTSD both showed improvement in depressive symptoms Women with PTSD had poorer functioning
Preferences for Depression Care African Americans and Latinos particularly likely to prefer psychotherapy over medication (Dwight-Johnson et al 2000) Compared to whites, African Americans: Less likely to find antidepressants acceptable Less likely to believe antidepressants are effective More likely to believe antidepressants are addictive (Cooper et al 2003)
What does it take to deliver effective depression care to disadvantaged persons?
The Chronic Care Model www.improvingchroniccare.org
Chronic Care Model components for depression 1. Community: linkages and partnerships with community agencies that provide psychosocial services 2. Health system: buy-in from leadership, translation into policies and procedures, resource allocation 3. Self-management support: education, activation for self-care and engagement in mental health care 4. Delivery system design: designated staff person for depression care, outreach and proactive follow-up 5. Decision support: treatment guidelines, training, access to consultation 6. Clinical information systems: disease registries to monitor outcomes and provide reminders
Isolated components not as effective Screening, or screening with provider feedback Dissemination of practice guidelines Provider education and training Referrals to mental health care settings Only 1/3-1/2 of patients follow-through Most never complete a course of treatment Quality cannot be assured Unutzer, Schoenbaum, Druss, et al: Transforming mental health care at the interface with general medicine: report for the presidents commission. Psychiatric Services 2006; 57:37-47.
IMPACT (Unutzer et al 2002) 1801 depressed older adults 12% African American 8% Latino 3% other minority groups 7 study sites Patients randomized to Collaborative Care vs. Usual Care
IMPACT Intervention 1. Care manager facilitates systematic treatment and follow-up a. Patient education / self-management support b. Supports antidepressant management prescribed by PCP c. Systematically tracks outcomes, side effects, treatment effectiveness (PHQ-9) d. Offers course of brief, evidence-based therapy (e.g., Behavioral Activation; Problem Solving Treatment) 2. Psychiatric consultation / caseload supervision 3. Stepped care: change treatment and increase intensity according to evidence-based algorithm if patient is not improving 4. Relapse prevention once patient is improved
IMPACT Effectiveness in Disadvantaged Populations 50 % or greater improvement in depression at 12 months 60% 54% 50% 40% 30% 20% 43% 19% 23% 42% 14% IMPACT Care Care as Usual 10% 0% White Black Latino Arean et al. Medical Care, 2005
Low-income patients also benefit Arean et al, Psychiatric Services, Aug 2007
Long-term outcomes 2 year follow up of IMPACT shows effect persists even 1 yr after program ends Compared to usual care, subjects have: >100 more depression free days Higher quality of life Decreased physical pain No significant difference in treatment cost between intervention and usual care
http://impact.son.washington.edu/
Partners in Care (Wells et al 2000) 46 primary care practices 1356 depressed patients 398 Latinos 93 African Americans 778 Whites Practices randomized to: Quality improvement for medication Quality improvement for psychotherapy Usual Care Interventions materials in English and Spanish developed by multicultural intervention team
Partners in Care Rates of appropriate depression care increased for all ethnic groups Rates of probable depression at 6 months: Latinos: 47% QI vs. 64% UC (p=0.02) African Americans: 24% QI vs. 56% UC (p<.01) Whites: 37% QI vs. 41% UC (NS)
Multi-faceted Oncology Depression Program 55 low-income Latinas with breast or cervical cancer randomized to collaborative care vs. usual oncology care MODP patients significantly more likely to have improvement in depression symptoms (OR=4.51 95% CI 1.07-18.93) Suggested effect on treatment adherence and survival Dwight-Johnson et al, Psychosomatics 2005
Adaptations to collaborative care for minority patients Increased outreach efforts Assistance with practical barriers to care Culturally adapted and translated: Educational materials Psychotherapy manuals Provider education should include attention to cultural sensitivity
Case Management improves adherence to CBT in Latinos 199 low-income primary care patients randomized to group CBT vs. group CBT + case management Supplemental case management associated with: Lower drop-out from therapy Greater improvement in depression and functioning outcomes among Spanish speakers Miranda et al, Psychiatr Serv 2002
Summary Rates of depression treatment are low among persons of color and the poor Although more studies are needed, current evidence suggests that available treatments are effective in persons of color Treatments may need to be adapted Collaborative care interventions are particularly effective for improving depression care for disadvantaged groups